Google Dials Back AI Overviews In Search Results, Study Finds

Last updated on

Recent research indicates that Google’s AI-generated summaries have undergone substantial changes since their initial introduction. A study conducted by SE Ranking, which examined 100,000 keywords, revealed that Google has significantly decreased the frequency of AI-generated summaries. However, when these summaries do appear, they are now more detailed compared to earlier versions.

The study delves into which subjects and industries are more prone to receiving AI-generated summaries. It also examines how these AI snippets interact with other search elements such as featured snippets and advertisements.

Here’s a summary of the findings and their implications for your SEO strategies.

Declining Frequency Of AI Overviews

In contrast to figures prior to the rollout, the study reveals that AI Overviews are now triggered in 8% of the analyzed searches, marking a 52% decline from levels observed in January.

Yevheniia Khromova, the author of the study, suggests that this decrease indicates Google’s more cautious strategy, aiming to enhance the accuracy and reliability of AI-generated responses. She states:

“The significant reduction in the presence of AI Overviews likely reflects Google’s efforts to increase the accuracy and credibility of AI-generated answers.”

Longer AI Overviews

Despite the reduced frequency of AI overviews, those that do appear now offer more detailed information. The average length of the text has increased by almost 25%, reaching approximately 4,342 characters.

Another significant change is that AI overviews now tend to link to fewer sources on average, typically containing just four links after expanding the snippet. However, it’s noteworthy that 84% of these overviews still include at least one domain from the top 10 organic search results for that specific query.

Niche Dynamics & Ranking Factors

The likelihood of encountering an AI overview differs significantly between various industries.

Industries such as relationships, food and beverages, and technology showed the highest propensity to prompt AI overviews.

In sensitive sectors like healthcare, legal issues, and news, the incidence of AI summaries appearing was minimal, less than 1%.

Search queries consisting of ten words or more were notably more inclined to trigger an AI overview, with a rate of 19%, suggesting that AI summaries are particularly beneficial for complex informational requirements.

Additionally, search terms with lower volumes and lower cost-per-click were more prone to featuring AI summaries.

Other Characteristics Of AI Overviews

Research indicates that 45% of AI overviews are accompanied by featured snippets, frequently sourced from the same domains.

Moreover, 87% of AI overviews currently appear alongside ads, up from 73% previously. This statistic suggests a heightened competition for advertising space in these contexts.

What Does This Mean?

SE Ranking’s research on AI overviews highlights several significant implications:

  1. Reduced Risk of Traffic Losses: With fewer searches triggering AI overviews that provide direct answers, there is a lower likelihood of organic listings being demoted or receiving decreased traffic.
  2. Most Impacted Niches: Industries such as relationships, food, and technology see a higher prevalence of AI overviews. Publishers operating in these sectors should closely monitor Google’s AI overview strategy to optimize their content strategies accordingly.
  3. Emphasis on Long-form and In-Depth Content: As AI snippets expand in length, businesses may need to focus on creating more comprehensive content that goes beyond what AI overviews typically cover. This ensures that they provide additional value and depth to users beyond the brief summaries generated by AI.

Looking Ahead

While there has been a recent decline in the number of AI overviews, it’s uncertain whether this trend will persist.

The evolution of AI overviews is expected to continue over time.

It is important to closely monitor developments, experiment with various approaches to addressing them, and adapt strategies accordingly.

Original news from SearchEngineJournal